Anzeige
Erschienen in:
2016 | OriginalPaper | Buchkapitel
Network Traffic Classification Model Based on MDL Criterion
verfasst von : Ying Zhao, Junjun Chen, Guohua You, Jian Teng
Erschienen in: Advanced Multimedia and Ubiquitous Engineering
Verlag: Springer Singapore
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Network traffic classification is elementary to network security and management. Recent research tends to apply machine learning techniques to flow statistical feature based classification methods. The Gaussian Mixture Model (GMM) based on the correlation of flows has exhibited superior classification performance. It also has several important advantages, such as robust to distributional assumptions and adaption to any cluster shape. However, the performance of GMM can be severely affected by the number of clusters. In this paper, we propose the minimum description length (MDL) criterion which can balance the accuracy and complexity of the classification model effectively by evaluating the optimal number of clusters. We establish a new classification model and analyze its performance. A large number of experiments are carried out on two real-world traffic data sets to validate the proposed approach. The results demonstrate the efficiency of our approach.